© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B SAEs
    5. Gemma-2B
    7. Residual Stream
    9. 10-RES-JB
    11. 185
    Prev
    Next
    INDEX
    Explanations

    references to the United States and its political structures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.10.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    chanind/openwebtext-gemma
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    Architecture
    standard
    Context Size
    1,024
    Dataset
    ctigges/openwebtext-gemma-1024-cl
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    50
    +0.32
    1.3%
    1978
    +0.07
    0.3%
    395
    +0.06
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1479
    +0.32
    0.07
    405
    +0.07
    0.06
    1310
    +0.06
    0.05
    Negative Logits
    <bos>
    -2.64
     intersper
    -0.79
     winked
    -0.76
     darted
    -0.71
     rehabilitate
    -0.68
     shuddered
    -0.67
     shivered
    -0.67
     dampen
    -0.66
     defray
    -0.65
     reappeared
    -0.65
    POSITIVE LOGITS
     soulign
    1.04
     véhic
    1.04
     Cfr
    0.98
     soggior
    0.95
     pleins
    0.94
     pican
    0.94
     Traité
    0.93
     rafra
    0.92
     Châ
    0.92
     considér
    0.89
    Activations Density 0.326%

    No Known Activations